What's new
What's new

Measuring lathe wear, or what amount of taper is "acceptable".

ichudov

Hot Rolled
Joined
Feb 1, 2008
Location
Illinois
I would like to know what amount of wear on a lathe is considered to be reasonable.

For example, if I cut a round bar between centers, say over 8", while using a good cutter and taking off little material, what amount of taper would be considered "decent". I understand that how much is acceptable depends on the intended use, so let me just say that the use for would be for hobby purposes.

This is a Clausing 6913 14x48 lathe.

I am about to try measuring its precision and wanted to know a little more.

i
 
I'm not crazy about much measurable taper on something only 8" long. I think you need to find out what the problem is if you are getting a lot of taper. Is the tail stock set over? Get a nice piece of stock an inch or two in diameter and about 12 or 15 inches long. Put a good center in it and take a few thousandths off and then measure it at both ends. On my lathe if I got .0001" difference I would be happy. Anything less and not so happy.
 
In "Tool and Manufacturing Engineers Handbook" third edition table 5-20, Standard Lathe Accuracy, there is a list of 28 tests and specifications for these machines.

In the 12-18 inch capacity column,
"Tail-stock taper alignment allows .0008 in a 12 inch test bar.
The same value is listed for"spindle taper run out, TIR at the end of a 12 inch test bar.
Vertical alignment of head and tail centers is 0- .003" (high at tail)
Chuck run out 3 inch out from chuck jaw on a bar the same diameter as the thru hole, 0-.003"



table 5-21, "Cutting accuracy", lists four tests.

roundness: mounted in chuck; 0.0004"
cylindrical turning: "on 20 inch long 2-7/8" dia "dumbell" (2-3/4" dia. between test surfaces. The two cut surfaces should be within 0.0015" chuck mounted, and 0.001
between centers.

Bar is low carbon steel, cut in three passes, roughing, shaping and test cut, using normal machining practices speeds and feeds

That is one reference.

CalG
 
Between centres will only measure tailstock alignment.


A more reliable definition of machine condition is a piece of bar clamped in the chuck with 6" out, with the tailstock end UNSUPPORTED and turned over it's length.
 
Actually, to qualify lathe bed wear I have used a long test bar and a test indicator.
Betwen centers to "adjust" the tail stock set over to -0-, then travel right down the length of the bar "mapping" the reading every inch or so. (do it a couple of times, and swap ends on the test bar etc, etc., then average the readings.

Now where did I leave that 3 foot test bar????? ;-)
 
Limy has the best and simplest test. Get the lathe bed level across the ways. Use a large diameter test bar and a sharp cutter to avoid deflection. This will catch bed wear near the chuck, where it's most commonly seen.

Depends how you define hobby work, but if you can hold a thousandth in this test, that will let you a large number of projects. From there you can tweak bed leveling and tailstock alignment to do at least as well for long turned-between-centers work.
 
Limy

If a jaw chuck is used and the chuck jaws are worn, bell mouth or sprung, "the cut a chunk and see" method is no better than the "tailstock set over test" method ;-) The wrong feature is being evaluated.

OH heck! Test and measurement. there is so much to know....
 
I finally gave up as the measurement equipment to tell the difference between bed wear, headstock alignment and various other things is beyond my resources. I leveled the bed to the best of my abilities, then shimmed the headstock until I could cut with minimum taper. On my 10" Logan, if I can bore a 2" long sleeve bearing with less than .0002" taper, I'm happy. Same thing with cutting a 2" long pin. Anything long I'll do between centers, so it's more dependent on the adjustment of the tailstock. It appears to take quite a lot of bed wear, maybe .005" or more vertical, to prevent achieving what I've described. Hey, if you want a long rod with perfect diameter, go visit a centerless grinding shop.

CH
 
If you aren't careful this sort of thing will drive you nuts searching for unattainable and irrelevant precision. Lathes are deliberately made to be, so far as is practical, quite tolerant of wear. For ordinary 12" to the foot scale work wear sufficient to seriously affect accuracy when using appropriate techniques is easily visible. Even with a fair bit of wear careful set-up and tweaking of adjustments will go a long way.

Best thing to do is to settle down with your thinking cap, a calculator (or computer) and some trigonometry assistance (Machinery's Handbook has what you need) to calculate what the effects of wear and maladjustment are on the tool path. Obviously the perfect, unworn tool path is parallel to the bed or, if you are worried about facing at 90 degrees across it. Do it for different diameters and be prepared for surprises! Also interesting to calculate the drop due to vee way wear.

Along the way you will find out why levelling to remove twist is desirable.

By the way simply sticking a bit of material in a three jaw chuck and turning away is not an appropriate technique for accurate work. Just because every thing comes out OK when things are in half decent condition does not make it an appropriate technique. It does make it an acceptable alternative to use when you know what the limitations are relative to the proper way. Everything in life is a compromise, but it sure helps if you know what's being compromised.

Clive
 
Some thoughts.

If it's a used / worn machine and your'e within 1/2 thou over 6" then I'd leave well alone UNLESS you know what youre doing.
 
"If a jaw chuck is used and the chuck jaws are worn, bell mouth or sprung, "the cut a chunk and see" method is no better than the "tailstock set over test" method ;-)"


This is not true.
Once you mount stock in a chuck, no mater how out of whack, once you make a cut, that cut is true to the spindle axis. A worn chuck does not matter for this test.

--Doozer
 
I have one of them, it is a nice machine but not terribly rigid on it's own feet.

You will have to level the sh!t out of it regularly if not bolted to the floor, if it's not on a concrete foundation just for this purpose level it even more often.

If you do not want to get pissed off about tapered parts for the rest of your life, invest $400.00 or so in a Mito or Starret level, you will save at least that much in medication over several years.

YMMV
 
"If a jaw chuck is used and the chuck jaws are worn, bell mouth or sprung, "the cut a chunk and see" method is no better than the "tailstock set over test" method ;-)"


This is not true.
Once you mount stock in a chuck, no mater how out of whack, once you make a cut, that cut is true to the spindle axis. A worn chuck does not matter for this test.

--Doozer

I think you're both right. Doozer's right that runout induced by the chuck is not an issue, provided the workpiece can be rigidly coupled to the spindle, but CalG is right if the chuck is so bell-mouthed that the work is not rigidly held at the outboard end of the jaws.

In this latter situation, the work will not spin in line with the spindle axis, so all alignment bets will be off. However this is easily remedied by inserting suitable shim stock offcuts in the outboard nip of each chuck jaw. (Not to remedy runout, but to prevent wobble relative to the jaws)

I agree with the recommendations to evaluate headstock alignment by chucking a short workpiece, preferably with two collars to eliminate the effects of tool wear.

Whereas if what matters is the ability to turn long workpieces to a consistent diameter (not so much in respect of taper, but barrel vs pincushion) then the relevant tests involve a long test bar - personally I prefer to carry out a static test using a straight, precision ground bar of substantial diameter in relation to length, held between centres, and a DTI held in the toolpost at centre height.

Adjust the tailstock so the DTI reading is the same at both ends, then take readings at regular intervals, then turn the bar (without turning the spindle) 180 degrees, repeat, and take the mean (at each measurement station) of the two readings.

This will quantify the net sum of bed wear and bed twist in the region you are interested in (assuming both are present, and at some level of measurement, both inevitably will be)
 
Troup has it!...... In all ways.
My thoughts regarding the chucking were in consideration of "looseness" Very likely the chuck and bedways arein similar condition....it just follows.
Carefull work can negate such defficiency.
 
Be careful too, I am not sure about that exact clausing, but many machines have adjustable headstocks too. Nardini, Some Clausings, Many Others.

I bought a machine about 5 years ago, it was a clausing from the 1970's. Got it for $100, because the guy bought it and was so fed up with the wear on it, he bought it on Ebay, it was shipped to him, and said he got ripped off on the condition. Well I had a machine that needed parts, so perfect match. Supposedly this guy was an expert machinist for 20 years.

I went to check this out, for $100 I was not gonna even look at it, was just gonna take it. I told him on the phone it was sold, never saw a picture or anything, just was a listing with phone number.
He said it was cutting a .010 taper in only about 3". I expected to see a mess. But it had the parts on it I needed, (thread dial and a few handles)

I got there and glanced at it, like I said for $100 I was taking it, and expecting to see a worn machine with a few parts I needed. I looked at it and thought the paint looked pretty darn good for a machine with so much wear. Then went to look closer, looked at the bed, and thought to myself, damn these ways look NICE. Oh well, still thought nothing of it. We picked it up with his lift, put it on my trailer and away I went.

Got back to the shop and started playing around with it, and looked closer and saw it was not worn at all. Looked at the headstock, noticed it looked like many of the new machines we carry, ADJUSTABLE.

Spent an hour getting it all set in, and had it turning within about .0005" over 6".
I called him to explain, told him what I did, he said it still was a POS and has to be worn, etc. I offered it back to him because I felt bad, and he did not want it. (I didnt say I got it within .0005, I just said much better than it was)
He was a pretty rude person too afterwards, a crab ass really. So I did not feel bad at all, especially since I called and told him and he said I didnt know what the hell I was talking about, maybe I need to learn to read a mic. Good for him, maybe he should learn to look into things instead of jumping to a conclusion and being hard headed.

Moral of the story, look closely, if its cutting a lot of taper, it may not be bed wear.
 
What an awesome story! I had several similar stories.

1. Diesel motor written off by military as "unrepairable scrap H7". It just needed seal replacement on the injection pump.
2. Military dental compressor "unrepairable scrap H7", had one tiny tube to a pressure switch that accidentally was disconnected.
3. Compressor with a "bad motor" that needed a new cap.
 
I'm with Samy. If it's fairly close, and you don't kow what you are doing, and don't have a nice level to get the bed right, don't screw with it.

Unless the headstock has been tweaked (by someone, or by a nasty meeting of chuck and cross slide or compound), I'd look for visible signs of wear on the bed. If none severe, get a piece of material nearly as long as your machine can handle (I prefer about 2" diam aluminum for a machine this size) and chuck it with a tailstock center in the outboard end. Take a .020 or so cut, just to true things up, then whiz it at about .005. and mic both ends. Adjust the tailstock to get it down to within a thou or so over this entire distance, and then go back and mic every two inches or so along the work. Use a vernier mic and record the readings in .0001" increments.

What you will get is sort of a map of error on your machine. If it has uniform taper from end to end, you are in exceptionally good shape and can get this taper to virtually nothing. More likely, you are going to see the work get a bit larger up near the chuck, where wear is present on the bed and dropping the tool. The work will be slightly football shaped. Nothing you can do about it, but you know it is there and can work around it.

If you want to check alignment from there, run the tailstock halfway up the bed and run another shorter test piece. If it matches your original map, you know the headstock is pointed down the bed. If it comes out with a larger taper, you can deduce which direction things are out of whack with some simple experimentation with the TS (moving side to side or shimming).

The POS Summit at work has about .008 football taper in 3ft, right in the middle of the bed. Having no other choice, I have marked rings on the work, .001, .002, .003, .004, .005, .006, .007, .008, -.007, -006, etc.. all the way back down to turn a shaft within .001 over that length. That's just how you work around it.
 
The POS Summit at work has about .008 football taper in 3ft, right in the middle of the bed. Having no other choice, I have marked rings on the work, .001, .002, .003, .004, .005, .006, .007, .008, -.007, -006, etc.. all the way back down to turn a shaft within .001 over that length. That's just how you work around it.

Think you an I went to the same school Mike :)
 
I think one of the first things to do is to chuck up the bar you intend to test cut, put a sensitive indicator on the end, and push it around with your finger. A two inch aluminum bar hanging out 6 or 8" deflects quite a lot under finger pressure, not to mention cutting pressure. Also the spindle bearings, spindle, and everything in between.

I learned this in the process of aligning the headstock on my last manual lathe. When you start adjusting things down to a few tenths or closer, you can really only do it for that particular size of that particular material with that particular cutter at that particular depth of cut - and probably on that particular day.
 
I think one of the first things to do is to chuck up the bar you intend to test cut, put a sensitive indicator on the end, and push it around with your finger. A two inch aluminum bar hanging out 6 or 8" deflects quite a lot under finger pressure, not to mention cutting pressure. Also the spindle bearings, spindle, and everything in between.

I learned this in the process of aligning the headstock on my last manual lathe. When you start adjusting things down to a few tenths or closer, you can really only do it for that particular size of that particular material with that particular cutter at that particular depth of cut - and probably on that particular day.

That is very good advice.

That is why I prefer to bore a hole to tweak the lathe levelling because the bar deflection is more or less constant, the piece of test pipe is relatively rigid because it is large diameter, and chucking on a piece of pipe using 4 jaw roughing jaws probably eliminates wobble of the part due to chuck wear.

And if the lathe tweakage for boring contradicts the tweakage for turning, so be it, I'd be tempted to leave it for boring straight. However, there is nothing stopping one from tweaking the levelling screws one way or the other, depending on the job of the day (if the machine is really bad).
 








 
Back
Top